The 1925 Brick Duplex With a Cold Second Floor
A two-story brick duplex built in 1925 has balloon framing with no fire blocking between floors. The second-floor bedrooms stay 6-8°F colder than the first floor all winter even after the owners replaced the furnace and sealed the visible attic floor with fiberglass batts. A blower door test reads 4,800 CFM50 on a 3,400 sq ft conditioned area. During a cold-day walkthrough you feel a steady draft at the baseboard of the second-floor exterior wall and see rust stains on the first-floor ceiling joists near the exterior wall. The attic has 4 inches of old cellulose and open soffits with no baffles. The clients want the upstairs warm and have a $9,000 budget.
1. Diagnose why the second floor stays cold, and rank the three most likely contributors with the physical mechanism for each.
Read the worked answer
The cold second floor is a stack-effect and thermal-boundary problem, not a furnace problem.
Balloon-framing wall cavities act as a chimney. In balloon framing, exterior wall stud bays run continuously from the sill to the eaves. Warm indoor air enters the cavity low, rises, and exits at the top. The draft at the second-floor baseboard is the supply end of that loop, and the rust on the first-floor ceiling joists near the wall marks where moist air has been condensing on cold framing. This is the dominant cause: it both cools the upstairs wall and dumps moisture into the assembly.
Missing attic air sealing and no soffit baffles. The 4-inch cellulose is thin (roughly R-14) and the open soffits let wind wash the insulation and connect the attic to the outside. The 4,800 CFM50 on 3,400 sq ft is roughly 1.4 ACH50-equivalent leakage, high for the envelope, and much of it is the top-of-wall bypass feeding the balloon cavity.
No fire blocking / thermal bypass at the floor line. The open stud bays also mean the second-floor band area is effectively uninsulated and connected to the first-floor ceiling cavity, so the floor platform itself is a cold bridge.
Ranking: balloon-cavity stack flow first (it explains the draft and the rust), attic bypasses second (they drive the stack), floor-line bridging third. Simply adding attic R-value without sealing would not fix the upstairs.
2. Given the $9,000 budget, propose a scope of work with rough cost allocation and state what you would defer.
Read the worked answer
Sequence by air sealing before insulation, and prioritize the stack.
- Dense-pack the exterior wall cavities (all sides, both floors) with a stabilized cellulose or fiberglass dense-pack. This blocks the vertical convection loop and adds R-value. Budget roughly $4,500-$5,500 for a 3,400 sq ft duplex with two stories of wall.
- Air seal the attic plane (top plates, penetrations, chimney chase) and install soffit baffles before adding insulation. Budget $800-$1,200.
- Top up attic insulation to R-49 only after sealing; budget $1,000-$1,500 for the additional depth.
- Install fire blocking in the balloon cavities at the floor line as part of the dense-pack work; this is a code and safety item and should be written into the scope, not deferred.
Defer: replacing windows, exterior wall re-siding, and any ductwork changes. The budget does not allow them, and they do not address the diagnosed mechanism. Also defer a second blower door until after the work, but plan and budget for it as the verification step. Warn the client that the rust indicates moisture, so a moisture check of the first-floor ceiling cavity should be done before closing up walls.
3. How would you verify the fix, and what metric would you report to the client?
Read the worked answer
Run a post-retrofit blower door test under the same conditions (same door setup, same outdoor temperature band, same baseline) and compare CFM50 and the normalized leakage. A drop from 4,800 CFM50 toward roughly 2,800-3,200 CFM50 would be a credible result for dense-packing plus attic sealing on this house. Also do a zone-pressure diagnostic (pressure pan or infrared scan of the second-floor wall) to confirm the balloon cavity is no longer communicating with the interior.
Report two numbers to the client: CFM50 before and after, and the estimated annual heating/cooling energy saved from the leakage reduction. Pair those with a comfort metric: the measured temperature difference between floors on a cold day, before and after. The temperature delta is what the client actually cares about, and it is the honest test of whether the stack loop was broken.